Academic employment of accounting doctoral graduates

全文

(1)c(.

(2) THE CHINESE ACCOUNTING REVIEW Vol. 25 PP.99·117 February 1991. ACADEl\fIC EMPLOYMENT OF .ACCOUNTING DOCTORAL GRADUATES. C. Jevons Lee. Tulane University. Mei-Hwa Lin. ~ational Cheng Chi University. Abstract. i. ~. 1\. I. I. This paper studies the pattern of academic employment of ac counting doctoral programs. The paper shows' how many doctoral graduates of top 40 accounting departments are e~ployed by the same class of institutions at three p~int in time, 1976, 1980, and 1985. Moreover, the paper examines the emplo~ment pattern in terms of major private and p~blic institutions. Qs.ing the ~Spearman correla tion coefficient, we show some characteristics of perceived quality that rr:tay be associated with the academic employment of doctoral gradu ates. We also find evidence of segmentation of academic accountants' labor market that is consistent with Nikolai and Bazley's coalition formation hypothesis.. 1. Academic Employment of Accounting Doctoral Graduates A dominating concern for a potential doctoral student is the quality of the doctoral program and the academic employment of the doctoral graduates. While many studies have been devoted to the measurement of a program quality in terms of various proxy variables such as survey opinion, number of publications and number of citations, little information has been 100. 4.

(3) ~. THE CHINESE ACCOUNTING REVIEW. 101. gathered about academic employment of doctoral graduates. The purpose of this paper is to provide information about the academic employment of t he doctoral graduates in the U.S. Moreover, the paper will examine the relationship between various characteristics of the perceived quality of doctoral programs and the academic employment of their graduates. The three frequent measures for institution quality are: 1. survey opinions (Carpenter et al., 1974);. 2. number of publications in major journals (Bazley and Nikolai, 1975;. Andrews and McKenzie, 1978; Jacobs et al., 1986); and. 3. number of citations (Brown ~ndGardner, 1985). Although all these measurements can affect an institution's reputation, they may not be equally correlated with the academic employment of their graduates. 1 It would be of great interest for prospective doctoral students as well as for program administrators to know what aspect of an institu tion's reputation is closely associated with the academic employment of the institution's doctoral graduates. The ability to place graduates in top accounting departments can itself be viewed as a measure of quality. To the extent that various proxy vari ables measure different aspects of quality and reputation, this paper shows how the academic employment of an institution's graduates complements other measures of quality in the literature. Moreover, Nikolai and .Bazley (1977) applied the theory of organiza tional interaction to show that coalition formation among accounting de . partments is a function of perceived prestige and geographiC location. They demonstrated that schools tend to hire graduates from institutions with . similar characteristics. Another purpose of this paper is to revisit Nikolai and Bazley's study with updated and expanded data.. 2. Ranking Accounting Departments 1 Two other approaches can be found in the literature . Koch, Merino, and Berman. (1984) analyzed the publications of accounting doctoral graduates. Snowball (1986) stud. ied the accounting department's academic strength in the area of accounting experiments. on human judgement.. ,I..

(4) ,i. 102. -t1tift£. Ranking institutional quality attracts continuous interest of students, educators, and administrators. In the literature,. the reputation of an ac counting department is sometimes measured by the quantity and quality of publications produced by the department's students and faculty. In other circumstances, a more direct measurement of reputation is adopted by sur veying the opinion of faculty and administrators. A brief literature review is given here to set the stage for our analysis. Carpenter, Crumbley, and Strawser (1974) ranked accounting faculties and doctoral programs according to the survey of accounting educators' opinions. Twenty-nine institutions were ranked in their study. Taking a dif f~rent approach, Bazley and Nikolai (1975) ranked accounting departments by counting the lines of authorship in the Accounting Review, Journal of Accounting Research, The Journal of Accountancy, and Management Ac.;. counting for a period from January 1968 through July 1974. They assigned publication credits to accounting departments according to three different criteria: one is based on the location of the publishing faculty member at the time of publication, the second is based on the location of the publish ing faculty member in.the 1974-75 academic year, and the third is based on the institution that granted the author's doctoral degree. Andrews and McKenzie (1978) adjusted Bazley and Nikolai's (1975) results by attaching a quality weight to each journal. In an effort to update Bazley and Nikolai's study, Jacobs, Hartgraves, and Beard (1986) evaluated the quality of accounting doctoral programs in terms of the number of articles published by their doctoral graduates in eight journals from January 1972 through December 1984. These journals include, in addition to the four journals in Bazl~y and Nikolai (1975), Financial Executive, Internal Auditor, Abacus, and'CPA Journal. Taking a fresh ap~roach, Brown and Gardner (1985) assessed the re search contribution of accounting faculty members and doctoral graduates according to the numr er of times their papers were cited. The 1976-1982 volumes of four journals, Accounting Organizations and Society, Journal of Accounting and Economics, Journal of Accounting Research, and The Accounting Review, were analyzed for the purpose of ranking accounting departments. In this paper, we examine the academic employment of doctoral gradu ates from various accounting programs. All accounting doctoral programs. 4.

(5) ". THE CHINESE ACCOUNTING REVIEW. 103 i i ,. I. are classified into" top" institutions and the" other" institutions. We define a top accounting department as the institution which is included in above rankings. All these top accounting departments have doctoral programs; they are the major demanders and suppliers in the labor market of aca demic accountants. The doctoral program that is not included in above rankings is grouped as the "other" institution.. 3. Data and Methodology Our study is composed of three basic elements: collecting the data about the faculty members who teach in top accounting departments, the estimation of a transition matrix and the estimation of marginal transition probabilities. We collect a sample of 40 top accounting departments from the four following articles: 1. The 29 accounting departments ranked in Table 2 of Carpenter, Crum. bley, and Strawser (1974, hereafter CCS). 2. The 15 accounting departments ranked in Bazley and Nikolai (1975, hereafter BN). 3. The 25 accounting departments ranked in Brown and Gardner (1985, hereafter BG). 4. The 25 accounting departments ranked in Table 2 of Jacobs, Hart graves, and Beard (1986, henceforth JHB). Names of faculty members teaching at these forty schools, their doctoral degrees, and the year of their graduation were obtained from Hasselback's Accounting Faculty Directory. Only individuals who have a doctorate with an accounting concentration were selected. 2 From these data, we can derive 2In Hasselback's Accounting Faculty Directory, a dot after the graduating year denotes a doctorate with an accounting concentration. Otherwise, it would be assumed that the person listed has a doctorate from other disciplines. Unfortunately, errors were found in these denotations, even though every effort was made to keep the information as correct as possible. Moreover, to improve its accuracy, we sent the preliminary report of this paper to the chairpersons of all forty ranked schools for data confirmation. However, the response rate is very low.. I. I.

(6) t-1Hf~. 104. a transition matrix G as in equation (1) XI,I. G=. xI,40. (1). Xi,i X40,l. X40,40. where xi,i is the number of graduates from school i teaching in the school j. The transition matrix illustrates how an object of observation moves from one state to another state of nature. For example, Leontief's input-output table is a transition matrix. The sum of each row in equation (1), Xi = Ei xi,i' is the n~mber of graduates froID school i teaching at top accounting departments. Let Yi be the total number of graduates from school i, then the percentage of gradu . ates from department i teaching in a particular accounting department j, Pi,i' would bexi,i / Yi. The value of Pi,i is an estimate of transition proba bility. Hence, we can calculate the marginal transition probability, Pi, from equation (2) Pi. = LPi,i = xi/y,. (2). The value of Pi indicates the probability that a graduate of school i will obtain a teaching position in one of top schools. The marginal transition probability indicates the chance for a subject (say, a student) to move from the initial state i to an ending state j. For example, the chance that a T University graduate will enter a medical school and the chance that a W University graduate will find a job in investment banking are marginal transition probabilities. The study in Nikolai and Bazley (1977) is an analysis of transition matrix G, whereas this study is an analysis·of marginal transition probabilities Pi.. 4. 'I he Placement Pattern Three volumes, 1986, 1981, and 1976, of Hasselback's Accounting Fac ulty Directory are analyzed. 3 In each volume, we calculate the transition matrix and transition probability for graduates from top 40 accounting 3 we. started this project in 1987. The la.test issue availa.ble then was 1986 (for the. 4.

(7) ". THE CHINESE ACCOUNTING REVIEW. 105. departments. Moreover, we group the graduates according to their gradu ating years by three time spans. First, we group all graduates within the past five years (e.g., 1981-85 graduates in the 1986 volume). Second, we group all graduates of past ten years (e.g., 1976-85 graduates in the 1986 volume). Third, we group all graduates who are still active in teaching (e.g., all graduates in the 1986 volume). The five-year and ten-year cut-offs for data grouping roughly match the career milestones of an average doctoral graduate. Persons who graduated within the most recent five years are generally at the assistant professor level, with a few exceptions of associate professor, and those who received their doctoral degree within the most recent ten years are mostly assistant professors and associate professors, with some exceptions of full professor. This matching is not perfect; it is adopted for convenience. The employment of a junior faculty member (a recent graduate within five years) may depend very much on the reputation of his/her alma mater and that of his/her dissertation advisors. The employment of a more senior faculty member may depend on training, drive, and other factors that have less to do with the reputation of his/her alma mater. It would be interesting to see whether the employment pattern of the recent graduates is different from those elder graduates. Moreover, faculty are known to move around., In so doing, at 'any point in time, they may not be at a top accounting department. Measuring the employment pattern over three different time horizons can alleviate the measurement error arising from faculty mobility. Table 1 summarizes the sample attributes. The first row shows that among the 2,405 doctorates who graduated from top 40 accounting depart ments before 1985, 51 teach at their own alma mater and 440 teach at one of top 40 accounting departments. Only 56 out of 496 (11%) faculty in top 40 accounting departments are "outsiders". Hence, in 1985 18.3% (440/2405) of graduates from top 40 schools teach at top 40 schools and they hold 89% (440/496) of the faculty positions in these schools, while only 6% (56/936) of graduates from other schools teach in top 40 schools. These results are consistent with the findings in Nikolai and Bazley (1977). academic year 1986-87}. We choose three time points of five years interval so that we can measure the employment pattern of the very recent graduates, "semi-established" graduates and "weU-established" graduates. Naturally, the choice of the five-year time interval is rather ad hoc..

(8) t-1tif~. 106 . TABLE 1 Sample SUmmary. A Period of Graduation Top Up to 1985 76-85 81-85. B Total # of Graduates from Top 40 Schools 2405 854 430. Up to 1980 71-80 76-80. 1975 982 424. Up to 1975 66-75 71-75 -_ _ Notes:. 1551 1009. .... ...... - - ...... ~55~ ~.~.. C. D. E. # of Graduates. # (%) of Graduates. # (%) of Graduates from. Teaching in Alma Mater 51 (2.12%) 18 (2.11%) 7 (1.63%). Teaching in Top 40 Schools 440 (18.3~~ 222 (26.0%) 129 (30.0%). Other Schools Teaching in Top 40 Schools 56 (6.0~) 46 (7.1%) 26 (7.3%). F Total # of Faculty in 40 Schools 496 268 155. 55 (2.78%) 20 (2.04%) 9 (2.12%). 401 (20.3%) 233 (23.1%) 141 (33.3%). 27 (4.7%) 24 (4.9%) 17 (5.8%). 428 257 158. 331 (21.3%) 202 (20.0%) !.3QJ2..l1. 3%L_ '-. 10 (3.5%) 10 (4.1%) 7 (3.6%). 341 212 137 --. L. .52 (3.35%) 17 (1.68%) !t!-43%1_ _. 1. Data are collected from 1986, 1981, and 1976 Hasselbaclr.'s Accounting Faculty Directory. 2. The period specified in column A includes both ends. For example, the doctoral studenta that graduated in 1976-85 include those graduated in 1976 and 1985. 3. Graduat!s in column D also include those in column C.. However, the statistics in the past 15 years also show an increase in both number and percentage of doctorates graduated from other schools teaching in top 40 schools. The third row shows that about 30% of recent doctoral graduates (during 1981- 85) from top 40 accounting departments have an appointment in top 40 accounting departments, while 7.3% of re cent doctoral graduates from other accounting departments get this type of position. Schultz, Meade and Khurana (1989) documented that the academic environment for accounting professors has changed over the period of 1961 1987; the average reported success rates for assistant professors seeking promotion and tenure decline precipitously. This paper shows that the academic market for doctoral graduates has changed in a different way. Comparing three 'Periods of graduations in Table 1 (1971-75, 1975-80, and 1981-85), we can see that the chance for a graduate from a top 40 school to find employment in a school of same class increased in the late 70s (from. ".

(9) ". THE CHINESE ACCOUNTING REVIEW. 107. 23.3% in 1971-75 to 33.3% in 1976-80) but becomes stagnant in the 1980s. On the other hand, the chance for a graduate from a other-than-top-40 school to find employment in a top 40 school steadily increased from 3.6% in 1971-75 to 7.3% in 1981-85. The ranking of doctoral programs in terms of percentage of graduates teaching at top 40 accounting departments, by different periods is shown in Table 2. The top five schools for the period up to 1985, 1980, and 1975 are the same, except for some changes in their relative rankings. These schools, Rochester, Carnegie- Mellon, Kansas, Stanford, and Cornell, have at least two features in common. First, these programs are all relatively small. Their sizes are presented in the last column of Table 3, which shows that Rochester granted seven Ph.D.'s up to 1985, Carnegie-Mellon 16, Kansas 13, Stanford 45, and Cornell 20. In contrast, the top two schools rated in Table 3, Illinois and Texas, have respectively 268 and 164 doctoral gradu ates. Second, all the five schools emphasize interdisciplinary research; there are no clear departmental boundaries in these schools. Columns (2) and (3) of Table 2 respectively present the percentage of doctoral graduates of the most recent 10 years and 5 years who are now teaching in top 40 schools. Hence, Column (2) reports the teaching location of a doctoral program's alumni who are either associate professors or as-' sistant professors, with some rare exceptions being full professors. Column (3) reports the teaching position of a program's alumni who are assistant professors, with some rare exceptions being associate professors. For a very small doctoral program, the results in Table 2 can be mis leading. For example, only one student graduated from the University of Pennsylvania in 1981-1985. Since he happens to teach at a top 40 sc~ool, he creates a" 100% batting average" for the University of Pennsylvania and gets it ranked in the first place. Table 3 ranks the "success" of a doctoral, program in terms of number of graduates teaching in top 40 accounting departments. The large accounting doctoral programs,' all in large state universities, occupy the top sev~n slots ranked in Table 3. The small programs offered at Cornell, Kansas, and Rochester, which are listed as the top five in Table 2, are ranked in the third quartile from the top in Table 3. Hence, rankings in terms of number and percentage give strikingly different perspectives. The.

(10) 1,' ~-~]"-. ;{1. ·::;;,1. 108. t-1Hf~. TABLE :I. Ranking by Percentage of Doctoral Graduates Employed in Top 40 Schools. University Rocheater (1912) Carnegie-Mellon (1951) Kansas (1970) Stanford (1939) Co!"nell (1968) Michigan (1939) Ohio State (1950) Chicago (1922) U. Penn (1913) U. Washington (1956) Berkeley (1929) Iowa (1951) Minnesota (1936) N. Carolina (1951) Northwestern (1956) VPI (1916) Purdue (1969) Texas (1934) SUNY·Buffalo (1951) Penn State (1961) Illinois (1939) Harvard (1935) Michigan State (1959) UCLA (1962) Wisconsin (1953) Oklahoma (1961) Florida (1956) Indiana (1950) Oregon (1964) Columbia (1952) Alabama (1953) Arizona (1910) Missouri (1941) Colorado (1966) Georgia (1910) Pittsburgh (1932) LSU (1943) NYU (1944) S. California (1963) MITl196~. Up to 1985 (1) 11 (1) 56 (2) 46 (3) H (4) 40 (5) 39 (6) 32 (1) 31 (8) 29 (9) 28 (10) 21 (11) 26 (12) 26 (12) 25 (14) 24 (15) 22 (16) 22 (16) 21 (18) 20 (19) 20 (19) 19 (21) 11 (22) 11 (23) 14 (24) 14 (24) 13 (26) 13 (26) 12 (28) 12 (28) 11 (30) 9 (31) 9 (31) 8 (33) 1 (34) 1 (34) 4 (36) 4 (36) 4 (36) 3 (39) o (4ql. Notes:. 1916-1985 (2) 61 (3) 60 (1) 61 (3) 69 (2) 21 (11) 62 (6) 44 (10) 54 (8) 61 (3) 38 (13) 31 (14) 24 (21) 40 (12) 26 (18) 29 (16) 22 (25) 0(39) 42 (11) 50 (9) 21 (26) 23 (22) 11 (1) 25 (19) 25 (19) 23 (22) 18 (29) 23 (22) 16 (30) 20 (21) 33 (21) 19 (28) 12 (31) 6 (36) 10 (32) 6 (36) 8 (35) 4 (38) 9 (33) 9 (33) 0(39]. 1981·1985 (~. 100 (1) 61 (6) 61 (6) 86 (3) 29 (23) 61 (6) 43 (11) 60 (9) 100 (1) 55 (10) 45 (15) 42 (18) 46 (14) 36 (20) 44 (16) 8 (31) 0(38) 50 (11) 15 (4) 20 (26) 26 (24) 15 (4) 50 (11) 50 (11) 20(26) 25 (25) 31 (19) 13 (32) 18 (28) 33 (21) 11 (29) 10 (33) 10 (33) 14 (SO) 0(38) 14 (30) 9 (35) 33 (21) 9 (35) 0(38). Up to 1980, Up to 1915 (4) (5) 100 (1) 100 (1) 62 (2) 50 (4) 60 (3) 11 (3) 45 (5) 44 (5) 54 (4) 80 (2) 43 (6) 31 (8) 42 (8) 41 (6) 35 (9) 30 (11) 16 (22) 0(35) 28 (13) 21 (11) 28 (10) 31 (10) 15 (24) 30 (11) 23 (16) 26 (14) 26 (12) 23 (18) 25 (13) 31 (8) 16 (22) 43 (6) 40 (1) 24 (15) 25 (16) 18 (18) 0(35) 25 (13) 24 (11) 22 (11) 26 (14) 15 (24) 13 (25) 11 (20) 20 (20) 11 (20) 20 (20) 11 (29) 10 (29) 15 (24) 6 (33) 13 (28) 15 (22) 14 (21) 12 (21) 1 (33) 0(35) 18 (18) 22 (19) 12 (29) 10 (29) 8 (32) 0(35) 1 (33) 14 (23) 5 (39) 13 (25) 11 (31) 1 (32) 6 (36) 8 (31) 1 (33) 12 (21) 6 (36) 14 (23) 6 (36) 4 (34) o (4()) o {351. 1. The year in the bracket behind each school is the year of first graduate.. 2. The unit for each entry is percentage. The number in the bracket behind each entry indicates a rank. 3. None of the MIT graduates is identified as an accounting Ph.D. in Hasselbaek's. 4. Column (1) is based on all doctoral graduates up to 1985, Column (2) is based on the doctoral graduates who acquifed degrees between 1916 and 1985 and so on so forth.. ".

(11) ". THE CHINESE ACCOUNTING REVIEW. 109. TABLE 3. Ranking by Number of Doctoral Graduates Employed in Top 40 Schools. University Illinois Texas Michigan Ohio State Michigan State Minnesota U. Washington Stanford Berkeley Chicago Harvard Indiana Penn State, N. Carolina Missouri Wisconsin Northwestern Camegi&-Mellon , Alabama Florida Iowa Cornell, Kansas UCLA Columbia Rochester LSU Oklahoma VPI Georgia Colorado Oregon SUNY-Buffalo' . South California NYU, Arizona Purdue U. Penn Pittsburgh MIT Notes: The unit for. Total Doctoral Up to 1985 1976-1985 1981-1985 Up to 1980 Up to 1975 up to 1985 (1) (2) (4) (5) (6) i3) 268 ~1) 52 ~1~ 50 ~ll 12 ,~1) 17 ~~l 51 Pl 35 (2) 20 (1) 5 (8) S8 (2) 29 (2) 164 (2) 30 (3) 13 (3) 8 (2) 28 (4) 21 (4) 77 (12) 29 (4) 12 (4) 6 (4) 32 (3) 26 (3) 90 (7) 21 (5) 8 (7) 7 (3) 19 (5) 22 (4) 18 (5) 20 (6) 8 (7) 6 (4) 15 (9) 78 (11) 15 (6) 20 (6) 10 (6) 72 (13) 5 (8) 17 (6) 13 (7) 20 (6) 11 (5) 6 (4) 17 (6) 45 (22) 13 (7) 16 (9) 7 (11) 5 (8) 13 (10) 12 (9) 58 (17) 16 (9) 6 (14) 3 (16) 16 (8) 51 (19) 12 (9) 15 (11) 5 (16) 3 (16) 13 (10) 10 (12) 85 (9) 12 (12) 5 (16) 2 (22) 12 (12) 9 (14) 103 (6) 12 (12) 8 (7) 4 (13) 10 (13) 5 (20) 60 (16) 12 (12) 8 (7) 5 (8) 48 (20) 4 (24) ,9(14) 11 (15) 4 (21) 3 (16) 8 (16) 145 (3) 12 (9) 11 (15) 7 (11) 4 (13) 81 (10) 7 (19) 5 (20) 10 (17) 5 (16) 4 (13) 41 (26) 8 (16) 9 (14) 9 (18) 6 (14) 2 (22) 3. (26) 16 (35) 8 (16) 8 (19) 4 (21) 2 (22) 9 (14) 7 (16) 88 (8) 8 (19) 7 (11) 6 (4) 6 (24) 5 (20) 64 (15) 8 (19) 5 (16) 5 (8) 3 (27) 3 (26) 31 (28) 8 (19) 4 (21) 2 (22) 7 (19) 4 (24) 20 (32) 6 (23) 4 (21) 2 (22) 13 (37) 6 (24) 5 (20) 6 (23) 3 (27) 1 (29) 42 (25) 7 (19) 6 (18) 5 (25) 5 (16) 3 (16) 7 (19) 7 (16) 47 (21) 5 (25) 4 (21) 2 (22) 5 (26) 7 (39) 1 (31) 4 (27) 1 (36) 1 (29) 7 (19) 10 (12) 105 (5) 4 (27) 3. (27) 3 (16) 3 (27) 1 (31) 32 (27) 4 (27) 4 (21) 1 (29) 0(35) 18 (34) 1 (35) 3 (30) 2 (31) 1 (31) 0(38) 45 (22) 3 (27) 3 (30) 2 (31) 1 (29) 3 (26) 2 (33) 44 (24) 3 (30) 3 (27) 2 (22) 0(35) 26 (29) 1 (35) 3 (30) 3 (27) 0(35) 15 (36) 3 (16) 2 (33) 2 (34) 2 (31) 1 (29) 2 (29) 3 (27) 65 (14) 2 (34) 1 (36) 1 (29) 6 (18) 3 (27) 54 (18) 2 (34) 2 (31) 0(35) 1 (29) 1 (35) 23 (31) 2 (34) 0(39) 2 (29) 0(38) 3. (27) 9 (38) 2 (34) 2 (31) 1 (29) 0(35) 7 (39) 1 (35) 1 (39) 1 (36) i (31) 1 (29) 25 (30) 1 (35) 0(40) o (39) 0(40) 0(35) 20 (32) 0138) each entry is number of graduates. The number in the bracket indicates a rank. I. I.

(12) t-itif~. 110 . large school looks "successful" in terms of number of graduates placed in top schools, whereas the small school look "successful" in terms of percentage of graduates placed in top schools.. 5. The Associations between Various Measures of Program Quality and Graguates' Employment To examine the association between graduates' employment pattern and various measures of perceived quality of doctoral programs, we calculated their rank correlations. Our results suggest two interesting implications. First, these associations may indicate relevant factors that account for the success of doctoral graduates who find faculty appointment in major re search universities. Second, to the extent that all these rankings measure certain aspects of quality, the associations among these rankings indicate their complementarity. Nine sets of rankings in the literature are presented in Table 4.4 ees is based on the opinion survey, BN, AM, and JHP are based on the number of author lines in publication count, and BG is based on the numbers of citations. BN assigned publication credits to accounting departments ac cording to three different criteria. The ranks of accounting departments in BN1 are calculated according to the location of the publishing faculty member at the time of publication, the ranks in BN2 are based on the lo cation of the publishing faculty member in the 1974-75 academic year, and the. ranks in BN3 are based on the institution that granted the author's doctoral degree (see respectively Tables 2, 3, and 4 of BN, 1975). AM1 and AM2 are respectively a weight-adjusted version of BN1 and BN2. BG1 and BG2 include different length of periods for counting the citation (see respectively Table 7c and 8c of BG). The references and the bases for the nine sets of rankings in the literature are summarized in footnote 1 of Table. 4. The notation # in the second column of Table 4 denotes the association between the rankings in Table 3 (which is based on the number of graduates 4We also ranked schools by their doctoral graduates teaching at top 10 and top 20 schools as listed in B01. The associations between these two rankings·and rankings in the literaturewere calculated. The results are similar to those correlations that are based on top 40 schools. therefore, only the results based on top 40 schools are presented.. ~:.".. 4.

(13) ". 111. THE CHINESE ACCOUNTING REVIEW. TABLE ... Spearman Correlations between Rankings by Academic Employment of Doctoral Graduates and Quality. Rankings in the Literature. '# or% '#. Literature CCS CCS BNl BNl BN2 BN2 BN3 BNS AMI AMI AM2 AM2 BGI BGI BG2 BG2 JHB JHB Notes:. %. '# %. '# %. '# %. '# %. '# %. '# %. '# %. '# %. Period up to 1975 up to 1975 up to 1975 up to 1975 up to 1975 up to 1975 up to 1975 up to 1975 up to 1976, up to 1975 up to 1975 up to 1975 up to 1985 up to 1985 up to 1985 up to 1985 up to 1985 up to 1985. Correlation 0.6176'" 0.390S* 0.1057 -0.1868 0.5615· -0.2552 0.7307*~. 0.3936 0.0969 -0.1429 0.3385 0.0198 0.3048 0.7099*· 0.2813 0.7341·· 0.5999·· ·0.0289. Period 1966-1975 1966-1975 1966-1975 1966-1975 1966-1975 1966-1975 - 1966-1975 1966-1975 1966-1975 1966-1975 1966-1975 1966-1975 1976-1985 1976-1985 1976-1985 197~1985. 1976-1985 1976-1985. Correlation 0.6141·* 0.4515· 0.1506 -0.2264 0.5156 -0.2926 0.5475· 0.3753 0.1329 ·0.1956 0.2619 -0.0374 0.1795 0.5188" 0.2136 0.5521" 0.4792· 0.0373. Period 1971·1975 1971·1975 1971-1975 1971-1975 1971·1975 1971-1975 1971-1975 1971-1975 1971-1975 1971-1975 1971-1975 1971-1975 1981-1985 1981-1985 1981-1985 1981-1985 1981-1985 1981-1985. Correlation 0.4789** 0.3647 0.0842 -0.0968 0.2981 -0.1122 0.4185 0.2111 0.0421 -0.1078 0.0889 0.0769 0.1169 0.4339· 0.0912 0.4722· 0.5252" 0.1219. 1. The reference!! and bases for the nine sets of rankings in the literature are summarized as follows:. ccs:. Carpenter et 11.1 (1974), based on the survey ofaccounting educators' opinions.. BN: Bazley and Nikolai (1975), based on the authorship in the Accounting Review, Journal of Accounting Research, Journal of Accountancy, and Management Accounting. BNI: based on the location of author at the time of publication. BN2: based on the location of author in 1974-75. BNS: based on the alma mater of the I!.uthor. AM: Andrew and McKenzie (1978), assigning different weight to the four journals in BN. AMI: a weight-adjusted version of BNl. AM2: a weight-adjusted version of BN2. BC: Brown and Gardner (1985), based on citation of faculty and graduates. BCI: based on all papers cited in four journals between 1976 and 1982. BC2: based on only those papers published in 1976-1982. JHB:Jacobe et aI (1986), based on the authorship of eight journ~ls during 1972-1984. 2. % indicl!.tes the relationship between Table 2 and the literature.. #-. indicates the relationship between Table S and the literature.. 3. The correlation is a Spearman Rank correlation, where • indicates significant at the 5% level and .. at the 1% level..

(14) ,:1. ~'r:~. t-lt"$f-1.t. 112. ".f. 1. 1. -_';\c. employed in top schools) and those in the literature, while the notation % indicates the association between the rankings in Table 2 (which is based on the percentage of graduates employed in top schools) and those in the literature. We tried to align the timing of our rankings with those in the literature for getting contemporaneous associations. For example, to match the rankings in CCS, which was published in 1974, we used the data in the 1976 volume of Hasselback's to calculate the correlation. Out of nine sets of rankings in the literature, Table 4 shows six sets of significant associations with the rankings in this paper. The opinion survey in CCS is significantly associated with the percentage as well as the num ber of doctoral graduates teaching in top 40 accounting departments. The numbers of author lines in BN2, BN3, and JHB are significantly associ-" ated with the number of doctoral graduates teaching in top 40 accounting departments. The number of citations in BG is significantly associated with the percentage of doctoral graduates teaching in top 40 schools. The . rankings in BNI, AMI and AM2 are not significantly associated with the employment opportunity of the doctoral program's graduates. To the ex tent that BN1, AMI, AM2 and our rankings measure certain aspects of academic quality, BNI, AMI and AM2 are complementary to our measure ments. The number of graduates placed in top 40 schools seems to be associated with the faculty opinion (CCS) and the quantity of publications (BN2, BN3, and JHB). However, Table 3 shows that the rankings in terms of number of graduates placed in top schools are positively correlated with program size. An accounting department that has more faculty and doctoral students tend to have larger qual!tity of publications, have larger number of related people expressing favorable opinion about the institution, and get more graduates placed in top schools than a smaller department. Hence, these number-based correlations in Table 4 are confounded by the size effect. The correlations based on the rankings in terms of percentage of gradu ate placement provide us a quite different perspective. Table 4 shows that the number of publications written by faculty members of an accounting department does not appear to be associated with the percentage of its· doctoral graduates obtaining employment in top schools. Rather, the qual ity of publications (measured by citation and faculty opinion) produced by an accounting department seems to be more closely associated with the employment opportunity of its doctoral"graduates.. 4. !. \. 1.

(15) ". THE CHINESE ACCOUNTING REVIEW. 113. 6. Placement to Public and Private Institutions The academic environment of public institutions is different from that of private institutions. Students who enroll in the 'doctoral program of a public institution may have preference to teach in a public institution. Stu dents who enroll in a private institution may prefer academic employment in a private institution. Moreover, Nikolai and Bazley's (1977) coalition formation hypothesis suggests that each institution has tendency to hire students from institutions of similar 'characteristics and mission. Tables 5 and 6 examine how the difference in education between public and private schools affect the pattern of academic employment. Table 5 reports a 3x3 transition matrix of students graduating from top private, top public and the other institutions who teach at top private or top public schools, or find employment other than teaching in top 40 insti tutions. The top public institutions produces the lion's share of doctoral graduates and also provides the largest number of faculty positions. The ratio of top-public/top-private graduates for all existing faculty at 1985 is about 4.5 to 1. The "market share" of the other schools increased drastically in the 1980s. Before 1975, the market share of top public schools was 68%, of top private schools was 18%, of the other schools was 15%. For those who graduated between 1981-85, the market share of top public schools declined to 47%, .that of top private schools declined to a mere 8%, while that of the' other schools jumped to 45%. This surge of new programs implies that various rankings of accounting departments will see significant change in t.he future. Table 6 provides the academic employment in terms of public/private school from each of top 40 institutions. The 3x3 transition matrices in Table 5 summarize these employment patterns. An independence test is applied to each matrix. Their chi-square statistics are. all significant at 1% level. Hence~ an institution tends to supply its doctoral graduates to the in stitutions with similar characteristics. For example, a top public institution tends to supply doctoral graduates to other top public institutions. Hence, there is a trace of market segmentation in the academic accountants' labor market. This result is again consistent with Nikolai and Bazley (1977)..

(16) ',;.t11 ~. 114. -tlt1f~. TABLE 5. The Employment Pattern. of Doctoral.Graduated of Top Private, Top Public and the Other Schools. Period Up to 1985 From:. To: Public Private Others* Total 1976-1985 Public From: Private Others Total 1981-1985 Public From: Private Others Total Up to 1980 Public From: Private Others Total Up to 1975 Public From: Private Others Total. Public Private OEM** Total 292 54 1620 1966 51 439 43 345 52 936 4 880 387 109 2845 3341 138 33 550 721 23 28 82 133 42 651 4 605 203 65 1237 1505 78 23 369 268 11 17 33 61 23 3 332 358 112 43 633 788 267 46 1284 1597 47 290 378 41 26 1 551 578 334 94 2125 2553 234 30 981 1245 39 239 28 306 275 10 285 0 69 272 1495 1836. X2. 179.91. 172.84. 113.69. 164.50. 136.45. Notes:. ,.. The term" others" indicates doctoral programs other than top 40 institutions.. **. "Public" indicates top public institutions, "Private" indicates top private institutior. and "OEM" indicates employment other than top 40 institutions .. •.

(17) TABLE 6. Academ i <: Employment of Graduates of Top 40 Schools:. Private Versus Public. ~;'niversity. Au!.bama Arizona Berkeley UCLA Carnegie-Mellon Chicago Colorado Columbia Cornell Florida Georgia Harvard Illinois Indiana Iowa. Kansas LS\!;,ic. .' MIT Michigan Michigan State Minnesota Missouri NYU. N. Carolina Northwestern Ohio State Oklahoma Oregon U. Penn Penn. State Pittsburgh. Purdue Rochester S. California Stanford SUNY-Buffalo Texas VPI U. Washington Wisconsin ,Total .I. Up to 1985 PRIVATE PUBLIC 8 2. 8 3 4 8. 1976-1985 PUBLIC PRIVATE 4 2 2. 5. 1. 2. 4 4 2 2 2 6 2. 2 2. 8 3 5 8 3 2 4 7 3 4. 11. 2. 44. 8. 13 5 5 3 1. 3 4 1. 12 7'. 1. 5" 3. 1. 25 19 15 11 10 7. 25 2 3. 1. 5 2. 5 2 2 3 4 2. 11 6 5 4 6 3 10 2 3. 2. 11 1. 1. 3 2. 1981-1985 PRIVATE PUBLIC 2 , 1 1. '.. 1. 3 1 1. 1. 5. 3 4. 1 8 2 5 2. 1. 4 1 1. 2 2 . 1 2 4. 1. 2 2 3. 6 5 3. 2 2 3. 3 1 2 2 2 1. 4 2 6 2 2. 2. 1. 4 1. 1. 1 1 11 1 31 4 17' 10 335. 7 5 3 5 2 1 5 4 3 4 42 11 3 5 6. 6 2 5 11. 24 19 13 8. 4. 6 2 2 9 8 1 1 1. 4 1 9 2 4 3 1 I 105 I. 1 5 1. 17 4 8 6 161. 4 1 6 2 3 2 1 61. 2. 1 4 1 4 4 89. 9 6. 14. 1. 3 1 3 7. 12. 1 4 6 19. 2 1. 1. 1. 9 1. 1. 8 5 25. 5 3 7. 1. 4 2. 14 2. 1. 34 1 13 7 308. 40. 3 2. 2. 2 2. 1. 4 3 4 2 5 1 1 45 9. 3 5. 10. 5 2 3 8. 5. 2 1. .. 7 4. 16 18. 3. 5. 1 2 1. 1 3. 2 7. 6. 4. 26. 3. 4. 12 5 262. 1. 3. 2. Up to 1976 PUBLIC PRIVATE 7. 1. 1. 2. 8. 1 1. Up to 1980 PUBLIC PRIVATE 1 8. 93. 69.

(18) t-~t1t~. 116. 7. Conclusion In this paper, we ranked the doctoral programs by the percentage and number of graduates teaching in top 40 accounting departments. The lead~ ing accounting programs in terms of percentage of graduates employed in top 40 accounting departments are generally small in size and highly in terdisciplinary. On the other hand, the leading programs in terms of the number of graduates employed in top 40 accounting departments are large in size, long in history, and departmentalized. We found that the number of faculty publications is not significantly associated with the probability of graduates securing teaching positions in top schools. Rather, the number of citations and the opinion of faculties are more closely associated with the employment pattern. Due to their sheer size, the large doctoral programs have more faculty publications and also have a larger number of graduates taking posts in top schools. We also find a trace of labor market segmentation in terms of a public/private grouping. Accounting departments in the U.S. tend to hire graduates from institutions with similar characteristics. The focus of various programs may change over time. Programs with out an initial research orientation may later have become research oriented. New programs must establish themselves and develop reputations. The in creasing market share of the other schools implies that many emerging doctoral programs may become the major research institutions in the fu ture. We find that the chance for a graduate of an institution outside of top 40 to find employment in· a top 40 school increased from 3.6% in 1971-75 to 7.3% in 1981~85.. References Abdel-khalik, A. R., and B. B. Ajinka. 1979. Empirical RE:search in Ac counting: A Methodological Viewpoint, Accounting Education Series #4, Sarasota (FL: American Accounting Association, 1979). Andrews, W. T., and P.B. Mckenzie. 1978. Leading Accounting Depart~ ments Revisited. The Accounting Review (January):135~138.. 4.

(19) •. THE CHINESE ACCOUNTING REVIEW. 117. Bazley, J. D., and L. A. Nikolai. 1975. A Comparison of Published Ac counting Research and Qualities of Accounting Faculty and Doctoral Programs. The Accounting Review {July):605-610. Brown, L. D., and J. C. Gardner. 1985. Applying Citation Analysis to Evaluate the Research Contributions of Accounting Faculty and Doctoral Program. The Accounting Review {April):262-277. Carpenter, C. G., L. D. Crumbley, and R. H. Strawser. 1974. A New Ranking of Accounting Faculties and Doctoral Program. The Journal of Accountancy {June):90-94. Hasselback, J. R., Accounting Faculty Directory, 1975-76, 1980- 81, 1986, 1987 Englewood Cliffs, NJ: Prentice-Hall,1976, 1981,1986, 1987 Jacobs, F. A., A. L. Hartgraves, and L. H. Beard. 1986. Publication Productivity of Doctoral Alumni: A Time-Adjusted Model. The Ac counting Review (January):179-187. Kida, T. E., and R. C. Mannino. 1980. Job Selection Criteria of Account ing Ph.D. Students and Faculty Members'. The Accounting Review (July):491-500. Koch, B., B. Merino and N. Berman. 1984. Publishing Patterns of Doc toral Graduates: A Preference for Perishing, in T. J. Burns, ed. Doc toral Programs in Accounting (The Ohio State University Press):351 366. Nikolai, L. A., and J. D. Bazley. 1977. An Analysis of the Organization Interaction of Accounting Departments. Academy of Management Journal {December):608-;-621. Schultz, J. J., J. A. Meade, and 1. Khurana. 1989. The Changing Roles of Teaching, Research, and Service in Promotion and Tenure Decisions for Accounting Faculty. Issue in Accounting Education {Spring):109 119. Snowball. D.1986. Accounting Laboratory Experiments on Human Judg ment: Some Characteristics and Influences. Accou~ting, Organiza tion, and Society {FebruarY):47-63..

(20)